Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/348209
Title: | Detection of normal and epileptic eeg signals using modified haar wavelet and neural network |
Researcher: | Vani S |
Guide(s): | Suresh G R |
Keywords: | Neurological disorders Epilepsy Engineering Biomedical |
University: | Anna University |
Completed Date: | 2020 |
Abstract: | Epilepsy is one of the neurological disorders of central nervous system which is characterized by periodic loss of consciousness with or without seizures related with unbalanced electrical activity within the brain. Recognition of epilepsy is time consuming process and also difficult. The Epilepsy seizers can be diagnosed only manually by skilled professionals from EEG recordings. The occurrence of epileptic seizure is unpredictable. The detection of seizer is done by calculating the spikes and fast waves in the EEG.The conventional techniques of analysis of EEG being tedious and time consuming, many automated epileptic EEG systems have appeared in recent years. Automated classification of EEG data is complicated by number of causes. The existence of epileptic form activity in the EEG validates the diagnosis of epilepsy which occasionally confused with other disorders producing similar seizure activity.In this study, an automated diagnostic model is designed and validated to support the physicians in improving the classification of epileptic seizures on EEG signals. To achieve this, the present study uses a modified Haar Wavelet Transform for decomposing the EEG recordings. It extracts useful features from the recordings and helps in reducing the complexity of computing the recordings. Therefore, the total number of additions and multiplications are reduced with increasing decomposition levels. The modified Haar wavelet transform is used for faster analysis of EEG signal. The signals are decomposed into sub bands and the coefficients are fed into a classifier. The present study further uses a machine learning approach namely Artificial Neural Network trained with Back Propagation algorithm to classify the epileptic seizures on EEG signals. newlineThe proposed approach is examined on various EEG dataset that includes and compared against different feature extraction methods. The newline newline |
Pagination: | xvii,128p. |
URI: | http://hdl.handle.net/10603/348209 |
Appears in Departments: | Faculty of Information and Communication Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf | Attached File | 85.08 kB | Adobe PDF | View/Open |
02_certificates.pdf | 438.8 kB | Adobe PDF | View/Open | |
03_vivaproceedings.pdf | 647.41 kB | Adobe PDF | View/Open | |
04_bonafidecertificate.pdf | 503.75 kB | Adobe PDF | View/Open | |
05_abstracts.pdf | 468.59 kB | Adobe PDF | View/Open | |
06_acknowledgements.pdf | 54.82 kB | Adobe PDF | View/Open | |
07_contents.pdf | 286.27 kB | Adobe PDF | View/Open | |
08_listoftables.pdf | 266.08 kB | Adobe PDF | View/Open | |
09_listoffigures.pdf | 655.53 kB | Adobe PDF | View/Open | |
10_listofabbreviations.pdf | 405.13 kB | Adobe PDF | View/Open | |
11_chapter1.pdf | 1.62 MB | Adobe PDF | View/Open | |
12_chapter2.pdf | 7.28 MB | Adobe PDF | View/Open | |
13_chapter3.pdf | 9.11 MB | Adobe PDF | View/Open | |
14_chapter4.pdf | 8.5 MB | Adobe PDF | View/Open | |
15_conclusion.pdf | 723.33 kB | Adobe PDF | View/Open | |
16_appendices.pdf | 540.37 kB | Adobe PDF | View/Open | |
17_references.pdf | 3.52 MB | Adobe PDF | View/Open | |
18_listofpublications.pdf | 94.36 kB | Adobe PDF | View/Open | |
80_recommendation.pdf | 664.41 kB | Adobe PDF | View/Open |
Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
Altmetric Badge: